How AI runs India’s quick commerce, ETBrandEquity

Image used for representative purpose (iStock) India has always been impatient. We want chai the second it boils, cricket scores before the ball lands, and now, groceries, fashion and beauty in under 30 minutes. This hunger for abhi—for now—has created a quick commerce industry worth INR 64,000 crore in FY25, projected to triple to INR 2 lakh crore by FY28, according to CareEdge. That’s 142 per cent CAGR between FY22 and FY25.

But behind the stopwatch lies a different race—the one algorithm runs every day to predict what we want, where we want it and how fast it can get to us without breaking the math.

A market running hot—and thin

The growth story is undeniable, but so is the strain. Blinkit reported 117 per cent year-on-year revenue growth in Q3 FY25, but also an INR 103 crore EBITDA loss. Swiggy Instamart posted INR 4,670 crore in gross order value in Q4 FY25, even as expansion costs widened losses. Quick commerce may win over consumers, yet the numbers still tell a harsher story.

Metros carry the weight. Dense order volumes make the math work, but in smaller cities the wheels wobble. Small towns drive 60 to 70 per cent of India’s overall retail demand, but account for just 20 per cent of quick commerce GMV (Redseer). A Delhi dark store can clear over 1,000 daily orders; a Lucknow outlet struggles at 600 to 700. Quick commerce remains an urban luxury.

The common code behind every cart

If groceries can be instant, why not fashion? That’s Myntra’s bet with M-Now, promising deliveries in as little as 30 minutes from 90,000 SKUs and more than 600 brands stocked across 40 dark stores. “Consumers now expect instant gratification not just for essentials, but also for trend-led and occasion-driven purchases,” says Lakshminarayan Swaminathan, vice president, product management and design at Myntra. Its predictive models decide which sneakers or kurtas to stock before anyone even clicks “buy.”

Nykaa has taken a similar leap in beauty, mapping demand for shades and SKUs at a hyper-local level. A lipstick that risks gathering dust in Jaipur can be instantly redirected to Delhi, while AI-powered analysers guide shoppers through hyper-personalised advice. Here, precision matters more than speed alone.

Amazon is collapsing decision time altogether. From StyleSnap’s image-based search to “Wear It With” suggestions, virtual try-ons, and a Skincare Analyser, the platform is designed to erase friction. “Generative AI will revolutionise customer interactions,” says Siddharth Bhagat, director, fashion and beauty at Amazon India. “Rufus, our AI assistant, answers questions, compares products and makes personalised recommendations.”

On the grocery front, Blinkit has leaned hard into AI for forecasting and routing. Its predictive models shine during festive spikes—Diwali or Rakhi, for instance—when the platform pre-stocks gifting SKUs to manage upwards of 500,000 orders a day. Swiggy Instamart takes a different tack, clustering orders and dynamically optimising routes in real time. Reports suggest its substitution logic—recommending replacements when an item is out of stock—is increasingly powered by machine learning.

Zepto, gearing up for its IPO, may be the boldest adopter. Its AI models factor in weather shifts, public holidays and even cricket match timings to anticipate surges. During Diwali 2024, it scaled to more than half a million daily orders while maintaining a 95 per cent + fill rate. For investors watching its unit economics, AI isn’t a feature—it’s the argument for scalability.

And then there’s BigBasket—the quiet but relentless engineer of the system. “While the early use of AI in enterprises was driven by marketing, its applications are now widespread and essential for efficiency and customer experience,” says Vipul Parekh, co-founder and chief marketing officer. What began with marketing automation has expanded across the value chain: AI-generated copy boosted productivity 20x; chatbots now draw on a unified database of reviews and emails to resolve complaints faster; search has become smarter, allowing queries like “organic staples under INR 200” or “total fatty matter in soaps” to return tailored results; hyper-local forecasting ensures dark stores stock exactly what their neighborhood needs; and logistics are optimised using real-time and historical traffic data. “From emails to route planning, AI is part of every workflow,” Parekh adds. “It’s no longer just an add-on—it’s the backbone.”

Representative image (iStock) Step back, and the pattern is clear. AI in quick commerce isn’t about one flashy feature. It’s spread across six critical zones:

Demand sensing: predicting what shoppers will want, and when

Assortment planning: deciding which SKUs to stock in which dark stores

Personalisation: tailoring recommendations and search results

Marketing automation: generating content, offers and nudges at scale

Customer support: chatbots, voice bots and sentiment analysis

Logistics: smarter inventory placement and last-mile delivery routing

The stopwatch may grab headlines. But the algorithm runs the business.

The next battlegrounds

AI can sharpen forecasts and optimise routes, but it can’t conjure density in smaller cities. As the Economic Times noted: “Quick commerce outside metros stretches margins and deepens gig-worker reliance.” For now, the model works best in India’s top metros.

Still, the next phase is already taking shape:

Occasion commerce: weddings, birthdays and office emergencies will drive premium, high-margin orders.

Conversational shopping: assistants like Rufus and Nykaa’s analysers will replace scrolling with chat-first discovery.

Margin discipline: subscriptions, tiered delivery fees and pruning weak stores will decide sustainability.

All under the watchful eye of investors. Zepto’s IPO could be the tipping point—does quick commerce become India’s retail future, or just a high-speed mirage?

The algorithm’s verdict

Quick commerce in India isn’t really about 30 minutes. It’s about collapsing the gap between wanting and having. AI is the invisible engine here, predicting cravings before they’re spoken, stocking shelves before the order comes in.

As Amazon’s Siddharth Bhagat says, “Generative AI will revolutionise customer interactions.” As BigBasket’s Vipul Parekh puts it, “AI is part of every workflow.” And as Myntra’s Lakshminarayan Swaminathan notes, “Consumers now expect instant gratification—not just for essentials, but for trend-led purchases too.”

The race isn’t against time, but against uncertainty. The winners won’t be the ones who shave off another two minutes—they’ll be the ones whose algorithms see around corners.

Because in India, where patience has never been our strong suit, the real prize isn’t speed. It’s certainty, delivered.>

• Published On Sep 1, 2025 at 08:09 AM IST

Join the community of 2M+ industry professionals.

Subscribe to Newsletter to get latest insights & analysis in your inbox.

All about ETBrandEquity industry right on your smartphone!

_{Images are for reference only.Images and contents gathered automatic from google or 3rd party sources.All rights on the images and contents are with their legal original owners.}